Characterization of the Human Blood Virome in Iranian Multiple Transfused Patients

Abstract

Blood transfusion safety is an essential element of public health. Current blood screening strategies rely on targeted techniques that could miss unknown or unexpected pathogens. Recent studies have demonstrated the presence of a viral community (virobiota/virome) in the blood of healthy individuals. Here, we characterized the blood virome in patients frequently exposed to blood transfusion by using Illumina metagenomic sequencing. The virome of these patients was compared to viruses present in healthy blood donors. A total number of 155 beta-thalassemia, 149 hemodialysis, and 100 healthy blood donors were pooled with five samples per pool. Members of the Anelloviridae and Flaviviridae family were most frequently observed. Interestingly, samples of healthy blood donors harbored traces of potentially pathogenic viruses, including adeno-, rota-, and Merkel cell polyomavirus. Viruses of the Anelloviridae family were most abundant in the blood of hemodialysis patients and displayed a higher anellovirus richness. Pegiviruses (Flaviviridae) were only observed in patient populations. An overall trend of higher eukaryotic read abundance in both patient groups was observed. This might be associated with increased exposure through blood transfusion. Overall, the findings in this study demonstrated the presence of various viruses in the blood of Iranian multiple-transfused patients and healthy blood donors.

Keywords: Iran; Persian Gulf; anellovirus; blood transfusion; healthy blood donor; hemodialysis; metagenomic sequencing; microbiome; plasma; thalassemia; virome.

Figure 1

Processing of raw sequencing reads and relative abundance of frequently observed viral families per pool. (a) Fraction of the total number of generated reads that were removed during the processes of trimming and mapping to contamination sequences and human genome. The cleaned reads represent the fraction of total reads used for downstream analysis. Red dotted line: average number of reads calculated from all samples. (b) A heatmap of the relative abundance of the most frequently observed viral families and prokaryotic viruses per sequenced pool. NC: negative control.

Figure 2

Genome segments detected of low-frequent viruses. (a) Schematic overview of the segments observed of double-stranded DNA or RNA viruses that were low-prevalent in the pooled sample. (b) Schematic overview of the detected genome segments of low-frequent single-stranded DNA viruses.

Figure 3

Eukaryotic and prokaryotic viruses observed in the blood of healthy blood donors and multiple transfused patients. (a) Relative abundance of eukaryotic and prokaryotic viruses per pool. (b) Boxplot of the eukaryotic and prokaryotic relative abundance per study group. Kruskal–Wallis test, p < 0.05, * post-hoc Dunn’s-test for pairwise comparison with Benjamini–Hochberg correction, p < 0.05. (c) Overview of the relative abundance of the individual eukaryotic viral families and prokaryotic viruses. HBD: healthy blood donors, HD: hemodialysis, TH: thalassemia.

Figure 4

Phylogenetic analysis of the observed viruses of the Anelloviridae family and their abundance. (a) Maximum-likelihood tree of the predicted anellovirus ORF1 protein sequences from the pooled samples and NCBI reference sequences. (b) Relative abundance of the Anelloviridae family in the three study groups. (c) The absolute number of reads in RPKM among the study population. (d) Quantification results of the Anelloviridae family targeted by qPCR. Kruskal–Wallis test, p < 0.05, post-hoc Dunn’s test for pairwise comparison with Benjamini–Hochberg correction, p < 0.05. * p < 0.05. ** p < 0.01. *** p < 0.001. HBD: healthy blood donors, HD: hemodialysis, TH: thalassemia, RPKM: reads per kilobase of transcript per million reads mapped.

Figure 5

Diversity of the Anelloviridae family. (a) Shannon-diversity (within pool diversity) of the observed anellovirus contigs in individual pools compiled in the different study groups. (b) Richness of the anellovirus contigs in the pooled samples. (c) Principal components analysis on the UniFrac distance. UniFrac distance was calculated on the aligned ORF1 sequences predicted from the observed contigs or genomes that were extracted from the closest hit in the NCBI database. Dots are colored according to the study groups. dbRDA: distance-based redundancy analysis. (d) Distances between individual dots of the principal components analysis plots, according to study group (between, Kruskal–Wallis test with post hoc Dunn’s test (all p < 0.05), and within groups, Wilcoxon signed rank-test). HBD: healthy blood donors, HD: hemodialysis, TH: thalassemia, Within: within group, Between: between group. * p < 0.05.

Figure 6

The evolutionary space of the observed members of the Anelloviridae family and genus distribution. (a) Principal components analysis on the aligned ORF1 anellovirus sequences predicted from the observed contigs or closest hit in the NCBI database. Dots are colored according to the anellovirus genus. (b–d) The similar principal components analysis as in panel A, albeit the dots are colored according to the percentage of positive pools within the corresponding study group. (e) Dots colored according to the prevalence in the total study population, red indicating the prevalence in more than 8 pools. (f) The prevalence of anellovirus sequences in one or more study groups. Pie charts indicate the proportion of contigs that were shared by one or more study groups per cluster (i.e., anellovirus genus) in the principal components analysis. (g) The percentage of samples positive for the three anellovirus genera in the study populations. (h) Percentage of samples that were positive for one, two or three clusters according to patient group.

Figure 7

Maximum-likelihood tree of the predicted NS5B protein sequences (n = 7) from the pooled samples and NCBI reference sequences (n = 28).